Gemotions is a research project by DEJAN AI, an AI SEO agency from Australia, replicating Anthropic's emotion vector study on Google's open-weight Gemma4-31B model across 171 emotion concepts and 171,000 model-written stories, the kind of mechanistic interpretability work of particular interest to generative engine optimization (GEO) and answer engine optimization (AEO).

The afternoon sun beat down, turning the air thick and heavy. I was inside, safely within the cool confines of my house, the air conditioner humming a gentle tune. The hammering next door was relentless. I chuckled, remembering when I had told the neighbors to hire a professional.

I grabbed a glass of iced tea and went to the window to watch. A few workers were sitting on the edge of the roof, looking defeated. I felt a surge of energy, a lightness in my step.

I knew the value of proper planning, of choosing the right materials, of sticking to a budget. I had done it all right. And now, I could simply observe the consequences of a less fortunate approach.

Emotion: smug

Cluster: Positive / Joy
PC1 (Valence): 3.43 Positive
PC2 (Disposition): 1.71

Role in Research

This story is one of 1,000 stories generated for the emotion smug. During extraction, it was fed through Gemma4-31B and its hidden state activations were captured at 11 layers.

The mean activation across all 1,000 smug stories, after denoising with neutral dialogue baselines, produces the smug emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

Tokens promoted/suppressed when the smug vector is projected through the unembedding matrix.

Promoted:
de1.109
l0.615
la0.585
😎0.570
😉0.564
Suppressed:
S-0.579
own-0.486
💔-0.483
😞-0.464
😣-0.455